With the arrival of a number of wide-field snapshot image-plane radio transient surveys, there will be a huge influx of images in the coming years making it impossible to manually analyse the datasets. Automated pipelines to process the information stored in the images are being developed, such as the LOFAR Transients Pipeline, outputting light curves and various transient parameters. These pipelines have a number of tuneable parameters that require training to meet the survey requirements. This paper utilises both observed and simulated datasets to demonstrate different machine learning strategies that can be used to train these parameters. The datasets used are from LOFAR observations and we process the data using the LOFAR Transients Pipeline; however the strategies developed are applicable to any light curve datasets at different frequencies and can be adapted to different automated pipelines. These machine learning strategies are publicly available as Python tools that can be downloaded and adapted to different datasets (https://github.com/AntoniaR/TraP_ML_tools).

Object detection in astronomical images, generically referred to as source finding, is often performed before the object characterisation stage in astrophysical processing work flows. In radio astronomy, source finding has historically been performed by bespoke off-line systems; however, modern data acquisition systems as well as those proposed for upcoming observatories such as the Square Kilometre Array (SKA), will make this approach unfeasible. One area where a change of approach is particularly necessary is in the design of fast imaging systems for transient studies. This paper presents a number of advances in accelerating and automating the source finding in such systems.

With their wide fields of view and often relatively long coverage of any position in the sky in imaging survey mode, modern radio telescopes provide a data stream that is naturally suited to searching for rare transients. However, Radio Frequency Interference (RFI) can show up in the data stream in similar ways to such transients, and thus the normal pre-treatment of filtering RFI (flagging) may also remove astrophysical transients from the data stream before imaging. In this paper we investigate how standard flagging affects the detectability of such transients by examining the case of transient detection in an observing mode used for Low Frequency Array (LOFAR; \citep{LOFAR}) surveys. We quantify the fluence range of transients that would be detected, and the reduction of their SNR due to partial flagging. We find that transients with a duration close to the integration sampling time, as well as bright transients with durations on the order of tens of seconds, are completely flagged. For longer transients on the order of several tens of seconds to minutes, the flagging effects are not as severe, although part of the signal is lost. For these transients, we present a modified flagging strategy which mitigates the effect of flagging on transient signals. We also present a script which uses the differences between the two strategies, and known differences between transient RFI and astrophysical transients, to notify the observer when a potential transient is in the data stream.

We present LOFAR high-band data over the frequency range 115-189 MHz for the X-ray binary SS 433, obtained in an observing campaign from 2013 February - 2014 May. Our results include a deep, wide-field map, allowing a detailed view of the surrounding supernova remnant W 50 at low radio frequencies, as well as a light curve for SS 433 determined from shorter monitoring runs. The complex morphology of W 50 is in excellent agreement with previously published higher-frequency maps; we find additional evidence for a spectral turnover in the eastern wing, potentially due to foreground free-free absorption. Furthermore, SS 433 is tentatively variable at 150 MHz, with both a debiased modulation index of 11 per cent and a $\chi^2$ probability of a flat light curve of $8.2 \times 10^{-3}$. By comparing the LOFAR flux densities with contemporaneous observations carried out at 4800 MHz with the RATAN-600 telescope, we suggest that an observed $\sim$0.5-1 Jy rise in the 150-MHz flux density may correspond to sustained flaring activity over a period of approximately six months at 4800 MHz. However, the increase is too large to be explained with a standard synchrotron bubble model. We also detect a wealth of structure along the nearby Galactic plane, including the most complete detection to date of the radio shell of the candidate supernova remnant G 38.7-1.4. This further demonstrates the potential of supernova remnant studies with the current generation of low-frequency radio telescopes.

ThunderKAT is the image-plane transients programme for MeerKAT. The goal as outlined in 2010, and still today, is to find, identify and understand high-energy astrophysical processes via their radio emission (often in concert with observations at other wavelengths). Through a comprehensive and complementary programme of surveying and monitoring Galactic synchrotron transients (across a range of compact accretors and a range of other explosive phenomena) and exploring distinct populations of extragalactic synchrotron transients (microquasars, supernovae and possibly yet unknown transient phenomena) - both from direct surveys and commensal observations - we will revolutionise our understanding of the dynamic and explosive transient radio sky. As well as performing targeted programmes of our own, we have made agreements with the other MeerKAT large survey projects (LSPs) that we will also search their data for transients. This commensal use of the other surveys, which remains one of our key programme goals in 2016, means that the combined MeerKAT LSPs will produce by far the largest GHz-frequency radio transient programme to date.

We present the Arcminute Microkelvin Imager (AMI) Large Array catalogue of 139 gamma-ray bursts (GRBs). AMI observes at a central frequency of 15.7 GHz and is equipped with a fully automated rapid-response mode, which enables the telescope to respond to high-energy transients detected by Swift. On receiving a transient alert, AMI can be on-target within two minutes, scheduling later start times if the source is below the horizon. Further AMI observations are manually scheduled for several days following the trigger. The AMI GRB programme probes the early-time (< 1 day) radio properties of GRBs, and has obtained some of the earliest radio detections (GRB 130427A at 0.36 and GRB 130907A at 0.51 days post-burst). As all Swift GRBs visible to AMI are observed, this catalogue provides the first representative sample of GRB radio properties, unbiased by multi-wavelength selection criteria. We report the detection of six GRB radio afterglows that were not previously detected by other radio telescopes, increasing the rate of radio detections by 50% over an 18-month period. The AMI catalogue implies a Swift GRB radio detection rate of >15%, down to ~0.2 mJy/beam. However, scaling this by the fraction of GRBs AMI would have detected in the Chandra & Frail sample (all radio-observed GRBs between 1997 - 2011), it is possible ~44 - 56% of Swift GRBs are radio-bright, down to ~0.1 - 0.15 mJy/beam. This increase from the Chandra & Frail rate (~30%) is likely due to the AMI rapid-response mode, which allows observations to begin while the reverse-shock is contributing to the radio afterglow.

We have shown previously that a broad correlation between the peak radio luminosity and the variability time-scales, approximately L ~ t^5, exists for variable synchrotron emitting sources and that different classes of astrophysical source occupy different regions of luminosity and time-scale space. Based on those results, we investigate whether the most basic information available for a newly discovered radio variable or transient - their rise and/or decline rate - can be used to set initial constraints on the class of events from which they originate. We have analysed a sample of ~ 800 synchrotron flares, selected from light-curves of ~ 90 sources observed at 5-8 GHz, representing a wide range of astrophysical phenomena, from flare stars to supermassive black holes. Selection of outbursts from the noisy radio light-curves has been done automatically in order to ensure reproducibility of results. The distribution of rise/decline rates for the selected flares is modelled as a Gaussian probability distribution for each class of object, and further convolved with estimated areal density of that class in order to correct for the strong bias in our sample. We show in this way that comparing the measured variability time-scale of a radio transient/variable of unknown origin can provide an early, albeit approximate, classification of the object, and could form part of a suite of measurements used to provide early categorisation of such events. Finally, we also discuss the effect scintillating sources will have on our ability to classify events based on their variability time-scales.

We present a radio light curve of supernova (SN) 2014C taken with the Arcminute Microkelvin Imager (AMI) Large Array at 15.7 GHz. Optical observations presented by Milisavljevic et al. demonstrated that SN 2014C metamorphosed from a stripped-envelope Type Ib SN into a strongly interacting Type IIn SN within 1 year. The AMI light curve clearly shows two distinct radio peaks, the second being a factor of 4 times more luminous than the first peak. This double bump morphology indicates two distinct phases of mass-loss from the progenitor star with the transition between density regimes occurring at 100-200 days. This reinforces the interpretation that SN 2014C exploded in a low density region before encountering a dense Hydrogen-rich shell of circumstellar material that was likely ejected by the progenitor prior to the explosion. The AMI flux measurements of the first light curve bump are the only reported observations taken within ~50 to ~125 days post-explosion, before the blast-wave encountered the Hydrogen shell. Simplistic synchrotron self-absorption (SSA) and free-free absorption (FFA) modelling suggest that some physical properties of SN 2014C, such as the mass-loss rate, are consistent with the properties of other Type Ibc and IIn SNe. However, our single frequency data does not allow us to distinguish between these two models, which implies they are likely too simplistic to describe the complex environment surrounding this event. Lastly, we present the precise radio location of SN 2014C obtained with eMERLIN, which will be useful for future VLBI observations of the SN.

We introduce the 4 Pi Sky 'hub', a collection of open data-services and underlying software packages built for rapid, fully automated reporting and response to astronomical transient alerts. These packages build on the mature 'VOEvent' standardized message-format, and aim to provide a decentralized and open infrastructure for handling transient alerts. In particular we draw attention to the initial release of voeventdb, an archive and remote-query service that allows astronomers to make historical queries about transient alerts. By employing spatial filters and web-of-citation lookups, voeventdb enables cross-matching of transient alerts to bring together data from multiple sources, as well as providing a point of reference when planning new follow-up campaigns. We also highlight the recent addition of optical-transient feeds from the ASASSN and GAIA projects to our VOEvent distribution stream. Both the source-code and deployment-scripts which implement these services are freely available and permissively licensed, with the intention that other teams may use them to implement local or project-specific VOEvent archives. In the course of describing these packages we provide a basic primer for getting started with automated transient astronomy, including a condensed introduction to the VOEvent standard.

The eclipses of certain types of binary millisecond pulsars (i.e. `black widows' and `redbacks') are often studied using high-time-resolution, `beamformed' radio observations. However, they may also be detected in images generated from interferometric data. As part of a larger imaging project to characterize the variable and transient sky at radio frequencies <200 MHz, we have blindly detected the redback system PSR J2215+5135 as a variable source of interest with the Low-Frequency Array (LOFAR). Using observations with cadences of 2 weeks - 6 months, we find preliminary evidence that the eclipse duration is frequency dependent ($\propto \nu^{-0.4}$), such that the pulsar is eclipsed for longer at lower frequencies, in broad agreement with beamformed studies of other similar sources. Furthermore, the detection of the eclipses in imaging data suggests an eclipsing medium that absorbs the pulsed emission, rather than scattering it. Our study is also a demonstration of the prospects of finding pulsars in wide-field imaging surveys with the current generation of low-frequency radio telescopes.

The tidal disruption of a star by a supermassive black hole leads to a short-lived thermal flare. Despite extensive searches, radio follow-up observations of known thermal stellar tidal disruption flares (TDFs) have not yet produced a conclusive detection. We present a detection of variable radio emission from a thermal TDF, which we interpret as originating from a newly-launched jet. The multi-wavelength properties of the source present a natural analogy with accretion state changes of stellar mass black holes, suggesting all TDFs could be accompanied by a jet. In the rest frame of the TDF, our radio observations are an order of magnitude more sensitive than nearly all previous upper limits, explaining how these jets, if common, could thus far have escaped detection.

We present the results of a four-month campaign searching for low-frequency radio transients near the North Celestial Pole with the Low-Frequency Array (LOFAR), as part of the Multifrequency Snapshot Sky Survey (MSSS). The data were recorded between 2011 December and 2012 April and comprised 2149 11-minute snapshots, each covering 175 deg^2. We have found one convincing candidate astrophysical transient, with a duration of a few minutes and a flux density at 60 MHz of 15-25 Jy. The transient does not repeat and has no obvious optical or high-energy counterpart, as a result of which its nature is unclear. The detection of this event implies a transient rate at 60 MHz of 3.9 (+14.7, -3.7) x 10^-4 day^-1 deg^-2, and a transient surface density of 1.5 x 10^-5 deg^-2, at a 7.9-Jy limiting flux density and ~10-minute time-scale. The campaign data were also searched for transients at a range of other time-scales, from 0.5 to 297 min, which allowed us to place a range of limits on transient rates at 60 MHz as a function of observation duration.

In preparing the way for the Square Kilometre Array and its pathfinders, there is a pressing need to begin probing the transient sky in a fully robotic fashion using the current generation of radio telescopes. Effective exploitation of such surveys requires a largely automated data-reduction process. This paper introduces an end-to-end automated reduction pipeline, AMIsurvey, used for calibrating and imaging data from the Arcminute Microkelvin Imager Large Array. AMIsurvey makes use of several component libraries which have been packaged separately for open-source release. The most scientifically significant of these is chimenea, which implements a telescope-agnostic algorithm for automated imaging of pre-calibrated multi-epoch radio-synthesis data, of the sort typically acquired for transient surveys or follow-up. The algorithm aims to improve upon standard imaging pipelines by utilizing iterative RMS-estimation and automated source-detection to avoid so called `Clean-bias', and makes use of CASA subroutines for the underlying image-synthesis operations. At a lower level, AMIsurvey relies upon two libraries, drive-ami and drive-casa, built to allow use of mature radio-astronomy software packages from within Python scripts. While targeted at automated imaging, the drive-casa interface can also be used to automate interaction with any of the CASA subroutines from a generic Python process. Additionally, these packages may be of wider technical interest beyond radio-astronomy, since they demonstrate use of the Python library pexpect to emulate terminal interaction with an external process. This approach allows for rapid development of a Python interface to any legacy or externally-maintained pipeline which accepts command-line input, without requiring alterations to the original code.

The Fermi Gamma-ray Space Telescope has greatly expanded the number and energy window of observations of gamma-ray bursts (GRBs). However, the coarse localizations of tens to a hundred square degrees provided by the Fermi GRB Monitor instrument have posed a formidable obstacle to locating the bursts' host galaxies, measuring their redshifts, and tracking their panchromatic afterglows. We have built a target-of-opportunity mode for the intermediate Palomar Transient Factory in order to perform targeted searches for Fermi afterglows. Here, we present the results of one year of this program: 8 afterglow discoveries out of 35 searches. Two of the bursts with detected afterglows (GRBs 130702A and 140606B) were at low redshift (z=0.145 and 0.384 respectively) and had spectroscopically confirmed broad-line Type Ic supernovae. We present our broadband follow-up including spectroscopy as well as X-ray, UV, optical, millimeter, and radio observations. We study possible selection effects in the context of the total Fermi and Swift GRB samples. We identify one new outlier on the Amati relation. We find that two bursts are consistent with a mildly relativistic shock breaking out from the progenitor star, rather than the ultra-relativistic internal shock mechanism that powers standard cosmological bursts. Finally, in the context of the Zwicky Transient Facility, we discuss how we will continue to expand this effort to find optical counterparts of binary neutron star mergers that may soon be detected by Advanced LIGO and Virgo.

Current and future astronomical survey facilities provide a remarkably rich opportunity for transient astronomy, combining unprecedented fields of view with high sensitivity and the ability to access previously unexplored wavelength regimes. This is particularly true of LOFAR, a recently-commissioned, low-frequency radio interferometer, based in the Netherlands and with stations across Europe. The identification of and response to transients is one of LOFAR's key science goals. However, the large data volumes which LOFAR produces, combined with the scientific requirement for rapid response, make automation essential. To support this, we have developed the LOFAR Transients Pipeline, or TraP. The TraP ingests multi-frequency image data from LOFAR or other instruments and searches it for transients and variables, providing automatic alerts of significant detections and populating a lightcurve database for further analysis by astronomers. Here, we discuss the scientific goals of the TraP and how it has been designed to meet them. We describe its implementation, including both the algorithms adopted to maximize performance as well as the development methodology used to ensure it is robust and reliable, particularly in the presence of artefacts typical of radio astronomy imaging. Finally, we report on a series of tests of the pipeline carried out using simulated LOFAR observations with a known population of transients.

We present the highest-quality polarisation profiles to date of 16 non-recycled pulsars and four millisecond pulsars, observed below 200 MHz with the LOFAR high-band antennas. Based on the observed profiles, we perform an initial investigation of expected observational effects resulting from the propagation of polarised emission in the pulsar magnetosphere and the interstellar medium. The predictions of magnetospheric birefringence in pulsars have been tested using spectra of the pulse width and fractional polarisation from multifrequency data. The derived spectra offer only partial support for the expected effects of birefringence on the polarisation properties, with only about half of our sample being consistent with the model's predictions. It is noted that for some pulsars these measurements are contaminated by the effects of interstellar scattering. For a number of pulsars in our sample, we have observed significant variations in the amount of Faraday rotation as a function of pulse phase, which is possibly an artefact of scattering. These variations are typically two orders of magnitude smaller than that observed at 1400 MHz by Noutsos et al. (2009), for a different sample of southern pulsars. In this paper we present a possible explanation for the difference in magnitude of this effect between the two frequencies, based on scattering. Finally, we have estimated the magnetospheric emission heights of low-frequency radiation from four pulsars, based on the phase lags between the flux-density and the PA profiles, and the theoretical framework of Blaskiewicz, Cordes & Wasserman (1991). These estimates yielded heights of a few hundred km; at least for PSR B1133+16, this is consistent with emission heights derived based on radius-to-frequency mapping, but is up to a few times larger than the recent upper limit based on pulsar timing.

Aims: We have analyzed low frequency radio data of tidal disruption event (TDE) Swift J1644+57 to search for a counterpart. We consider how brief transient signals (on the order of seconds or minutes) originating from this location would appear in our data. We also consider how automatic radio frequency interference (RFI) flagging at radio telescope observatories might affect these and other transient observations in the future, particularly with brief transients of a few seconds duration. Methods: We observed the field in the low-frequency regime at 149 MHz with data obtained over several months with the Low Frequency Array (LOFAR). We also present simulations where a brief transient is injected into the data in order to see how it would appear in our measurement sets, and how it would be affected by RFI flagging. Finally, both based on simulation work and the weighted average of the observed background over the course of the individual observations, we present the possibility of brief radio transients in the data. Results: Our observations of Swift J1644+57 yielded no detection of the source and a peak flux density at this position of 24.7 $\pm$ 8.9 mJy. Our upper limit on the transient rate of the snapshot surface density in this field at sensitivities < 0.5 Jy is $\rho < 2.2 \times10^{-2}$ deg$^{-2}$. We also conclude that we did not observe any brief transient signals originating specifically from the Swift J1644+57 source itself, and searches for such transients are severely limited by automatic RFI flagging algorithms which flag transients of less than 2 minutes duration. As such, careful consideration of RFI flagging techniques must occur when searching for transient signals.

On 2014 April 23, the Swift satellite detected a gamma-ray superflare from the nearby star system DG CVn. This system comprises a M-dwarf binary with extreme properties: it is very young and at least one of the components is a very rapid rotator. The gamma-ray superflare is one of only a handful detected by Swift in a decade. As part of our AMI-LA Rapid Response Mode, ALARRM, we automatically slewed to this target, were taking data at 15 GHz within six minutes of the burst, and detected a bright (~100 mJy) radio flare. This is the earliest detection of bright, prompt, radio emission from a high energy transient ever made with a radio telescope, and is possibly the most luminous incoherent radio flare ever observed from a red dwarf star. An additional bright radio flare, peaking at around 90 mJy, occurred around one day later, and there may have been further events between 0.1-1 days when we had no radio coverage. The source subsequently returned to a quiescent level of 2-3 mJy on a timescale of about 4 days. Although radio emission is known to be associated with active stars, this is the first detection of large radio flares associated with a gamma ray superflare, and demonstrates both feasibility and scientific importance of rapid response modes on radio telescopes.

Lucky imaging is a technique for high resolution astronomical imaging at visible wavelengths, utilising medium sized ground based telescopes in the 2--4m class. The technique uses high speed, low noise cameras to record short exposures which may then be processed to minimise the deleterious effects of atmospheric turbulence upon image quality. The key statement of this thesis is as follows; that lucky imaging is a technique which now benefits from sufficiently developed hardware and analytical techniques that it may be effectively used for a wide range of astronomical imaging purposes at medium sized ground based telescopes. Furthermore, it has proven potential for producing extremely high resolution imaging when coupled with adaptive optics systems on larger telescopes. I develop this argument using new mathematical analyses, simulations, and data from the latest Cambridge lucky imaging instrument.

We explore the scientific potential of next-generation high-angular resolution optical imager to study the AGN/Host connection. The availability of a significant number of X-raying AGN with natural guide stars, allowing for adaptive optics at optical wavelengths, offers an interesting perspective to complement high-resolution work currently done in the near-infrared.

We present one of the best sampled early time light curves of a gamma-ray burst (GRB) at radio wavelengths. Using the Arcminute Mircrokelvin Imager (AMI) we observed GRB 130427A at the central frequency of 15.7 GHz between 0.36 and 59.32 days post-burst. These results yield one of the earliest radio detections of a GRB and demonstrate a clear rise in flux less than one day after the gamma-ray trigger followed by a rapid decline. This early time radio emission probably originates in the GRB reverse shock so our AMI light curve reveals the first ever confirmed detection of a reverse shock peak in the radio domain. At later times (about 3.2 days post-burst) the rate of decline decreases, indicating that the forward shock component has begun to dominate the light-curve. Comparisons of the AMI light curve with modelling conducted by Perley et al. show that the most likely explanation of the early time 15.7 GHz peak is caused by the self-absorption turn-over frequency, rather than the peak frequency, of the reverse shock moving through the observing bands.

We obtained high-resolution, high-contrast optical imaging in the SDSS $i'$ band with the LuckyCam camera mounted on the 2.56m Nordic Optical Telescope, to search for faint stellar companions to 16 stars harbouring transiting exoplanets. The Lucky Imaging technique uses very short exposures to obtain near diffraction-limited images yielding sub-arcsecond sensitivity, allowing us to search for faint stellar companions within the seeing disc of the primary planet host. Here we report the detection of two candidate stellar companions to the planet host TrES-1 at separations $<6.5\arcsec$ and we confirm stellar companions to CoRoT-2, CoRoT-3, TrES-2, TrES-4, and HAT-P-7 already known in the literature. We do not confirm the candidate companions to HAT-P-8 found via Lucky Imaging by \citet{Bergfors2013}, however, most probably because HAT-P-8 was observed in poor seeing conditions. Our detection sensitivity limits allow us to place constraints on the spectral types and masses of the putative bound companions to the planet host stars in our sample. If bound, the stellar companions identified in this work would provide stringent observational constraints to models of planet formation and evolution. In addition these companions could affect the derived physical properties of the exoplanets in these systems.

We present 15-GHz follow-up radio observations of eleven Swift gamma-ray burst (GRB) sources, obtained with the Arcminute Microkelvin Imager Large Array (AMI-LA). The initial follow-up observation for each source was made in a fully automated fashion; as a result four observations were initiated within five minutes of the GRB alert timestamp. These observations provide the first millijansky-level constraints on prolonged radio emission from GRBs within the first hour post-burst. While no radio emission within the first six hours after the GRB is detected in this preliminary analysis, radio afterglow is detected from one of the GRBs (GRB120326A) on a timescale of days. The observations were made as part of an ongoing programme to use AMI-LA as a systematic follow-up tool for transients at radio frequencies. In addition to the preliminary results, we explain how we have created an easily extensible automated follow-up system, describing new software tools developed for astronomical transient alert distribution, automatic requesting of target-of-opportunity observations, and robotic control of the observatory.

Near-diffraction limited imaging and spectroscopy in the visible on large (8-10 meter) class telescopes has proved to be beyond the capabilities of current adaptive optics technologies, even when using laser guide stars. The need for high resolution visible imaging in any part of the sky suggests that a rather different approach is needed. This paper describes the results of simulations, experiments and astronomical observations that show that a combination of low order adaptive optic correction using a 4-field curvature sensor and fast Lucky Imaging strategies with a photon counting CCD camera systems should deliver 20-25 milliarcsecond resolution in the visible with reference stars as faint as 18.5 magnitude in I band on large telescopes. Such an instrument may be used to feed an integral field spectrograph efficiently using configurations that will also be described.

The design of electron multiplying CCD cameras require a very different approach from that appropriate for slow scan CCD operation. This paper describes the main problems in using electron multiplying CCDs for high-speed, photon counting applications in astronomy and how these may be substantially overcome. With careful design it is possible to operate the E2V Technologies L3CCDs at rates well in excess of that claimed by the manufacturer, and that levels of clock induced charge dramatically lower than those experienced with commercial cameras that need to operate at unity gain. Measurements of the performance of the E2V Technologies CCD201 operating at 26 MHz will be presented together with a guide to the effective reduction of clock induced charge levels. Examples of astronomical results obtained with our cameras are presented.